Semiconductor devices, in particular field-effect controlled switching devices such as a Metal Oxide Semiconductor Field Effect Transistor (MOSFET) and an Insulated Gate Bipolar Transistor (IGBT) have been used for various applications including but not limited to use as switches in power supplies and power converters, electric cars, air-conditioners, and even stereo systems.
Particularly with regard to power applications, semiconductor devices are often optimized with respect to low on-state resistance Ron at low chip area A, in particular a low product of Ron times A, fast switching and/or low switching losses. Furthermore, the semiconductor devices are often to be protected against high voltage peaks that may occur during switching of e.g. inductive loads.
DMOSFETs (double-diffused metal-oxide semiconductor field effect transistors) with channel structures manufactured using a double diffusion process for forming a body region and a source region of opposite doping type are often used, in particular in power circuits operating with large currents and/or at high voltages. So far, DMOSFETs are either implemented as planar DMOSFETs, i.e. DMOSFETs with a planar gate electrode structure, and trench-DMOSFETs in which the insulated gate electrodes are formed in trenches extending into the semiconductor substrate. Planar DMOSFETs require a comparatively large chip area A at given Ron, and are thus comparatively expensive. This applies in particular to planar MOSFETs with rated breakdown voltages above 30 V. As the MOS-channels of trench MOSFETs (T-MOSFETs) are designed along the typically vertical walls of the trenches, the cell pitch of the trench-DMOSFETs can be made small resulting in a comparatively small chip area A at given Ron. However, manufacturing is typically more complex for T-MOSFETs than for planar MOSFETs. Typically, the reduced chip area of T-MOSFETs outweighs the higher processing costs. However, energy-limited products, for example in automotive applications, and/or so-called multi-chip products requiring further signal pads and wiring may not fully benefit from the reduced required chip area of the T-MOSFET-structures because a certain chip area is required for energy dissipation during commutating and/or for the signal pads and/or for further wiring. This increases the costs of the products.
For these and other reasons there is a need for the present invention.